Skip to content

Qualcomm AI Engine Direct - Broadcast the constant base of quantized pow #24431

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
msluszniak wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Qualcomm AI Engine Direct - Broadcast the constant base of quantized pow #24431

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 8 additions & 0 deletions backends/qualcomm/builders/op_pow.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -12,6 +12,7 @@
from .node_visitor import NodeVisitor
from .node_visitor_manager import register_node_visitor
from .qnn_constants import OpElementWisePower, QNN_OP_PACKAGE_NAME_QTI_AISW
from .utils import is_parameter


# pow.Tensor_Scalar should fall in this visitor because LiftConstantScalarOperands pass
Expand Down Expand Up @@ -40,6 +41,13 @@ def define_node(
# tensor input
input_node = self.get_node(node.args[0])
input_tensor = self.get_tensor(input_node, node)
# pow.Scalar's base is lifted to a one-element constant, and quantized
# HTP ElementWisePower does not broadcast its first input: only the
# first output element is computed. Materialize the base at full shape.
if is_parameter(input_node, self.edge_program) and (
input_tensor.shape != out_tensor.shape
):
input_tensor = input_tensor.expand(out_tensor.shape).contiguous()

tensor_type = PyQnnManager.Qnn_TensorType_t.QNN_TENSOR_TYPE_NATIVE

Expand Down
Loading